What is SAGA?
SAGA is a fully autonomous research agent that works across your entire dataset. Simply provide it with a brief, and SAGA plans the investigation itself, queries your raw data, detects key narratives, pulls relevant charts, and delivers a complete, evidence-based report, all while you get on with something else.
Before You Begin
SAGA analyses data from your existing TRAC searches. Before creating a story, make sure you've created a TRAC search containing the data you'd like SAGA to analyse.
If you haven't created a search yet, head to TRAC first and build your search before returning to SAGA.
Here's a short video walkthrough of SAGA, which explains how it works and showcases the steps of creating a story to generating your first report.
How to Create Your First Story
Step 1: Open SAGA
From your Launchpad homepage, click SAGA to open the tool.
You'll be taken to the New Story page, where you can begin your analysis.
Step 2: Choose How to Start
There are two ways to create a story:
Write your own brief by entering a custom prompt describing the analysis you'd like SAGA to perform.
Use a template to get started quickly. Available templates include:
Crisis Signals
Brand Health
Competitive Gap
Audience Deep Dive
Trend Detection
Campaign Debrief
Templates provide a helpful starting point, but you can edit or expand the prompt to suit your specific requirements.
Note: Template prompts contain placeholder text inside [square brackets]. Be sure to replace these with the relevant brand, topic, industry, campaign, or audience before generating your story.
Step 3: Select Your TRAC Search
Select the TRAC search you'd like SAGA to analyse using the Search dropdown.
SAGA will use this search as the foundation for its analysis, so make sure you've selected the correct dataset before continuing.
💡Tip: To analyse multiple searches, leave the Search dropdown unselected and specify the search names you'd like SAGA to use directly in your brief.
Step 4: Choose a Timeframe
Next, select the date range for your analysis.
You can choose from:
Last 7 days
Last 30 days
Last 90 days
Custom date range
Step 5: Review and Refine the Approach
Before the analysis begins, SAGA presents a proposed research approach.
Take a moment to review it. SAGA may pose questions for you to answer too.
You can continue chatting with SAGA to refine the approach before running the analysis. For example, you could ask SAGA to:
Compare additional competitors
Focus on a particular audience
Investigate another market
Include specific charts or visualisations
Add or remove research questions
Narrow or broaden the investigation
Step 6: Run the Analysis
Once you're happy with the proposed scope, click Run Analysis.
SAGA will begin analysing your TRAC dataset and generating your report. Depending on the size of your dataset, the analysis may take a few minutes to complete.
Step 7: Continue Exploring Your Report
💡 Tip: The report SAGA first produces doesn't have to be the final output. Continue refining it by chatting back and forth with SAGA until it answers the questions or delivers the visuals that matter most.
Once the report has been generated, you can continue chatting with SAGA to explore your findings in more detail. For example, you can ask SAGA to:
Generate additional charts
Compare another time period
Break down the results by platform
Explore a specific narrative
Identify influential authors
Explain why a trend occurred
The more you interact with SAGA, the deeper your analysis becomes.
Starting a New Story
To begin a fresh analysis at any time, click New Story in the top-left corner of the screen.
This will take you back to the New Story page, where you can:
Select a different search.
Choose a template.
Write a completely new custom brief.
Generate another report.